st: ARCH and GARCH models with a t distribution

[Kit Baum <baum@bc.edu>]

sacrificial arch

Nick said

I surmise that you want t error terms, not Gaussian
(normal).

If so, this might require you
to clone the existing Stata commands and make
appropriate modifications. That is a guess, but if
so it would not be trivial. In addition, you would
have to decide whether the df of the Student t
distribution was a parameter to be estimated.



It is commonplace in other statistical packages to specify the  
assumed distribution of the errors in ARCH/GARCH models. For  
univariate ARCH/GARCH models (the only kind Stata's -arch-  
estimates), other software offers the choice of Normal, student t or  
GED (generalized error distribution) for the error distribution. In  
the latter two cases, an ancillary parameter is estimated (df for the  
t, 'shape parameter' for the GED, whose tails can be either  
platykurtic or leptokurtic). One might also want to fix the ancillary  
parameter.

Kit

Kit Baum, Boston College Economics and DIW Berlin
http://ideas.repec.org/e/pba1.html
An Introduction to Modern Econometrics Using Stata:
http://www.stata-press.com/books/imeus.html

*
*   For searches and help try:
*   http://www.stata.com/support/faqs/res/findit.html
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/